The Method and Theory of Site Catchment Analysis: A ReviewAuthor(s): Donna C. RoperSource: Advances in Archaeological Method and Theory, Vol. 2 (1979), pp. 119-140Published by: SpringerStable URL: http://www.jstor.org/stable/20170144 .Accessed: 28/03/2011 00:45

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=springer. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

Springer is collaborating with JSTOR to digitize, preserve and extend access to Advances in ArchaeologicalMethod and Theory.

http://www.jstor.org

4

The Method and Theory of Site

Catchment Analysis: A Review

DONNA C. ROPER

The spatial distribution of cultural phenomena is a basic theme in

archaeology. The culture area concept (Kroeber 1939), the concept of

horizon (Willey and Phillips 1958:33), and the notion of a settlement

pattern (Willey 1953:1) are but three ways in which archaeologists have

ordered space. The current interest of archaeologists in the description and explanation of site location using methods drawn from geography and

related disciplines can perhaps be viewed as simply the latest variation on

this basic theme. The recent publication of two books on the topic (Hod der and Orton 1976; Clarke 1977b) and the inclusion of two papers on

various aspects of locational analysis in this volume should certainly be

seen as some indicator of the status of locational analysis in archaeology in the second half of the 1970s.

Even cursory examination of the locational analysis literature (or one of

the recent syntheses of this literature, see especially Hodder and Orton

1976; or Hodder 1977) reveals a broad diversity in approach and tech

nique. One characterization of this literature would roughly distinguish two sets of approaches. The first set emphasizes the importance of

man-man relationships in structuring a community's ordering of space. Central place theory, the rank-size rule, and gravity models, among

others, would be included in this set of locational approaches. Three such

models are discussed by Crumley elsewhere in this volume.

Locational approaches in the second group assume the primacy of

119

ADVANCES IN ARCHAEOLOGICAL METHOD AND THEORY, VOL. 2

Copyright ? 1979 by Academic Press, Inc. All rights of reproduction in any form reserved.

ISBN 0-12-003102-7

120 DONNA C. ROPER

man-land relationships in determining site locations. Site catchment

analysis belongs with this latter group. It offers an alternative approach to

models based on central place theory (for example) in that it shows less concern with band spacing and population density, etc. (e.g., Wilmsen

1973; Schiffer 1975), as determinants of site location, and instead em

phasizes such considerations as the availability, abundance, spacing, and

seasonality of plant, animal, and mineral resources as important in deter

mining site location. However, it is distinguished from other man-land

approaches by the assessment of those resources within a demarcated area surrounding a site. That is, sites are conceived of as points at the

focus of an area throughout which economic activities were performed. The characteristics of this entire area, not just the immediate locus of the

site, are considered in inferring locational processes. It is essentially this basic distinction and the assumptions on which it is

based that unites the various studies that have been termed site catchment

analysis. Examination of the literature reveals a rather wide diversity of

purpose, scope, and technique used in site catchment analysis. It is the

purpose of this chapter to review the basis of site catchment analysis, the

techniques employed in its implementation, and the kinds of uses to which

it has been put.

WHAT IS SITE CATCHMENT ANALYSIS?

In proposing the term site catchment analysis, Vita-Finzi and Higgs

(1970:5) defined it as "the study of the relationships between technology and those natural resources lying within economic range of individual sites." The term catchment is drawn from the literature of geomorphology

where it is synonomous with drainage basin or watershed and denotes the area from which a stream draws its water. Similarly, the catchment of an

archaeological site is that area from which a site (or more properly, the

inhabitants of a site) derived its resources (see Vita-Finzi 1969a: 106?

which seems to be the first, albeit indirect, application of the term to the

archaeological case). Unlike drainage basins, the size, shape, and location

of a site's catchment may not be known in advance, frequently making it

necessary to use an initial estimate or approximation of the catchment

based on principles of settlement and land use, or some other knowledge. It is assumed that, in general, the farther one moves from an inhabited

locus, the greater the amount of energy that must be expended for pro curement of resources. Therefore, as one moves away from that locus, it

is assumed that the intensity of exploitation of the surrounding territory decreases, eventually reaching a point beyond which exploitation is un

SITE CATCHMENT ANALYSIS 121

profitable. Support for this assumption by users of site catchment analysis is drawn largely from the observation by Lee (1969:61) that the !Kung do

not normally go more than 6 miles (10 km) from their camps to procure resources, and the studies cited by Chisholm (1968:131) suggesting that

agriculturalists do not normally go even this far to tend their fields. Other

references could also be cited to illustrate the same point. It is further

assumed that prehistoric peoples were aware of this decrease in cost/

benefit ratio and located sites, moved their locations, and generally played out a settlement strategy that minimized the ratio of energy expended to

energy procured. It is further assumed that the site's inhabitants were willing to pay a

higher price (that is, expend more energy) for some resources than they were for others. Some resources, such as water, are so basic and so vital

that the distance to obtain them must be minimized; others are less

immediate, are "worth" more, and may therefore be gathered from

farther away. Because of this "hierarchy of importance of resources"

(Jochim 1976:54; see also Clarke 1968:506; Chisholm 1968:102-104), a

differentiation of use, or zonation, of the territory surrounding a settle

ment occurred. Both Jochim (1976:55) and Flannery (1976a: 117) have

described this situation in the archaeological literature?Jochim in theory, and Flannery empirically?and Chisholm (1968:101-110) used this as

sumption to quantify the relative cost of settling at alternative locations.

The biophysical environment is not uniform, however, either spatially or seasonally. The size, shape, and location of an individual site's catch

ment are therefore largely a function of the zonation, spacing, and sea

sonal differentials of resource zones exploited from the site. Decisions as

to whether or not the community should move to exploit seasonal differ

entials more economically will also be influenced by the structure of the

environment. Although many studies could be cited to support this as

sumption, the most dramatic example is to be found in Steward's (1938)

analysis, "Basin-Plateau Aboriginal Socio-Political Groups." Careful

consideration and contrasting of Steward's descriptions of the environ

mental zonation and settlement patterns of Great Basin tribes well illus

trates how major differences in environmental zonation correlate with

very different settlement strategies. Site catchment analysis was origi

nally developed as a response to the realization that at different times or

places the biophysical environment may offer very different possibilities for exploitation, given that there is a finite distance people are willing to

travel to exploit their environment. It is therefore a basic premise of site catchment analysis that site function and site location are correlated, and

that inferences can be made about function from knowledge of location.

The rationale for site catchment analysis is therefore relatively simple.

122 DONNA C. ROPER

It assumes no more than that human beings are refuging animals, rhyth

mically dispersing from and returning to a central place (Hamilton and

Watt 1970:263), differentially using a seasonally and spatially variable

landscape in a manner that generally is conservative of energy, but con

servative relative to a relative scale of values placed on needs and wants.

Although site catchment analysis relies largely on anthropological ob

servations, such as those cited above, for its theoretical justification, the

basic argument is not entirely without precedent in the literature of

economic geography. Both J. H. von Th?nen and A. Weber considered

location relative to availibility of resources, although in opposing fash

ions. Von Th?nen was concerned with general land use patterns that

developed around an isolated place and the balance between costs and

returns accruing to performance of various activities at a given distance

from this place. Weber reversed the procedure and directed his analysis toward finding an optimum location, given resource distribution and mode

of production (Clarke 1977a:21-23; Chisholm 1968:20-41 for discussion

and comparison of von Th?nen and Weber's work).

HOW IS SITE CATCHMENT ANALYSIS DONE?

In broad outline, site catchment analysis delimits a territory or set of

concentric territories surrounding a site and assesses the resource poten tial contained within that area. The territory assessed is that postulated to

be the area from which the greatest quantity of resources was derived.

Higgs et al. (1967), in their study of Paleolithic sites in Epirus, Greece, were the first to apply this form of analysis (although in a general form) when they sought to interpret the function of Kastritsa and its position in

the Advanced Paleolithic settlement system of Greece. They recognized the unequal seasonal and spatial distribution of resources, including the

animals represented by archaeological remains, and postulated seasonal

movements by the human populations that exploited these animals. Their

analysis then attempted to ascertain to what extent the Kastritsa site

would have been satisfactory for year-round occupation (Higgs et al.

1967:13). Although major interest was with interpretation of Kastritsa, its

location was also compared and contrasted with several other nearby sites. Summer and winter hunting conditions in Epirus were recon

structed, and rings 10 km in radius were drawn around each site. Re

sources within these 10-km radii were not quantified, but visual impres sions from inspection of maps conveyed the very different economic

potentials of the sites. Advanced Paleolithic sites of Epirus were then

SITE CATCHMENT ANALYSIS 123

grouped, based, in part, on inferences from seasonal potential (Higgs et

al. 1967:18). The study by Vita-Finzi and Higgs (1970) of Upper Paleolithic and

Neolithic sites in Palestine was actually the first to use the term site

catchment analysis, but, more important, it was also the first to discuss

some of the assumptions and principles of the type of analysis proposed. Vita-Finzi and Higgs (1970) were interested in evaluating whether their

sites necessarily represented sedentary economies, or whether trans

humance may have been practiced. The analysis worked from the premise "that attempts to solve this problem must take into account not only artifacts but also the possibilities inherent in the site situations them

selves" (Vita-Finzi and Higgs 1970:4). The "exploitation territory" of a site was defined as "the territory

surrounding the site which is exploited habitually" (Vita-Finzi and Higgs 1970:7). This territory was then used as an analytic device for examining the resources immediately accessible to a site's inhabitants.

Higgs et al. 's (1967) use of a 10-km radius had not accounted for sharp terrain differentials and thus did not fully account for energy expenditure differentials necessary to procure resources from different places within

the immediate vicinity of the site. Vita-Finzi and Higgs (1970) therefore

substituted time contours for circular radii, using 2-hour walks from a site

for hunter-gatherers, and 1-hour walks for agriculturalists. Using per

centages of land types within these time contours, they analyzed the

potential for occupation of their sites by comparison of the resource

potential of the exploitation territories of the sites. The result was what

they called "a very speculative possibility of a pattern of seasonal move

ment" (Vita-Finzi and Higgs 1970:22-26). In essence, it was a very

general, yet testable, hypothesis about seasonal movements of Natufian

populations. The Vita-Finzi and Higgs study therefore established site

catchment analysis as an inductive method for derivation of hypotheses about settlement system morphology.

The studies of Higgs et al. (1967) and Vita-Finzi and Higgs (1970)

exemplify the two techniques most commonly used for delimiting the

territory to be examined in a site catchment analysis?namely, the use of

circular territories of fixed radii and the use of time contours. Both means

of determining the area to be studied have since been widely used.

Walking 1 hour from agricultural sites and 2 hours from nonagricultural sites has been used by a number of European historians (Webley 1972;

Barker 1972, 1973, 1975b; Jarman and Webley 1975; Davidson 1976; Jarman 1976), although circles of fixed radii are more commonly used,

especially by Americanists but also by Europeans (Barker 1975a; Fagan

124 DONNA C. ROPER

1976; Moore et al. 1975; Noy et al. 1973; Clark 1972; Higgs and Webley 1971; Ellison and Harriss 1972; Clarke 1972; Dennell and Webley 1975;

Rossman 1976; Zarky 1976; Roper 1974, 1975; Peebles 1978). Possibly one

reason for the predominance of the latter technique is the fact that circles are more readily employed when data are taken from available maps rather than collected from actual walks taken in the field.

The uncritical acceptance of the Lee and Chisholm distance figures and

the mechanical use of these figures expressed as circular radii or time

contours are major problems with site catchment analysis as it is currently

practiced. Relieving this sterility in approach can perhaps best be initiated

by clarifying the catchment concept and explicitly stating its behavioral

referrent. Very early in the development of site catchment analysis the

terms territory and catchment were distinguished?the former as the area

immediately accessible to, a site's inhabitants, which was habitually

exploited, the latter as the total area from which the contents of a site

were derived (Higgs 1975:/jc). A territory as defined therefore became an

analytic device whose size was determined by ethnographic analogy. A

catchment, on the other hand, became a behavioral unit whose referrent

must be inferred from comparative knowledge of site territories, resource

distribution, site contents, and settlement system morphology. Obvi

ously, the better the analytic device approximates the behavioral unit, the

better the analysis of the catchment itself. Much of the site catchment

analysis literature has, however, tended to confuse and merge the two

terms. Thus, 2-hour or 10-km (or whatever) territories have been treated

as if they were actual catchments; time or distance contours as if the site's

inhabitants were on a 10-km-long leash. Several studies have attempted to

deal with this problem, however, and it is worth describing them briefly. Findlow and DeAtley (1974:4-5) explicitly noted the problem with

catchment approximation ("most catchment-hinterland types of analyses have failed to produce either a theoretical basis or empirical data to

support the use of some particular catchment size or shape") and have

attempted to resolve it. Their analysis of sites in the Animas Valley of

New Mexico formulated two site types and examined spacing along and

across drainages and between sites of the same type as well as different

types of sites* (Findlow and DeAtley 1974:38-40). Radiocarbon and

obsidian hydration dates were used to demonstrate contemporaneity of

sites. Observed spacings were taken as an estimate of the size and shape of catchments of different types of sites "as a preliminary step to examin

ing the relative uses and placement of sites within each catchment"

(Findlow and DeAtley 1974:54). Browman (1976) similarly calculated the

linear spacing of sites to interpret catchment size in Peru.

Cassels (1972b) attempted to determine the actual field (the term he

SITE CATCHMENT ANALYSIS 125

uses in place of catchment) of sites in the Waikato area of New Zealand by

constructing Thiessen polygons (Haggett 1965:247-248) around each site.

His assumption was that "the most likely boundary between the two sites

is a line equidistant between them" (Cassels 1972b:215). Such a means of

determining catchments assumes that all sites were contemporary, an

assumption of which Cassels is aware and which he accepts unless the

contrary is proved (Cassels 1972b:216). Strangely enough, however, once

he determined the size of the polygons, he used a set of concentric circles

to evaluate resource content and merely presented a frequency distribu

tion of size of polygons. Dennell and Webley (1975:102) eliminated overlaps of territories, prob

ably by a similar technique, and examined spacing. They too, however, used complete circles (2 km) to evaluate resources (Dennell and Webley 1975:105). Rossman (1976) and Brumfiel (1976) both truncated the over

lapping territories by drawing a straight line between the points of intersec

tion of the circles drawn around their sites. In both cases, the truncated

territories were used to evaluate resources.

Linear spacing (or some other measure of spacing) and Thiessen poly gons are both realistic approaches to estimation of catchment size and

shape. Their utility, however, is limited by several considerations. First,

they do assume contemporaneity of sites, and unless one can demonstrate

this to be so, the results could be highly misleading. Second, they assume a comprehensive listing of the sites whose spacing is being examined. This

could be a problem if analysis is being done of sites in an area that has not

been systematically surveyed or where a survey was done using quadrats, transects, or some other technique yielding an areally discontinuous sam

ple of sites. Third, use of either approach assumes no overlap of actually

exploited area, no trade, and no importation of resources?that is, it

assumes that the area within the polygon is the sole area exploited. The problem is, therefore, how do we approximate catchment size and

shape with anything but time or distance contours when the site sample is

nonsystematic, areally discontinuous, or noncontemporaneous? So far, no one has approached this problem directly. Flannery (1976a) dispensed with analytic devices altogether and attempted to determine site catch

ments empirically by starting with empirical data on plant, animal, and

mineral resources and asking from how far away they must have come

(Flannery 1976a: 103). It would seem reasonable that some data should be

available for most regions as to what resources were utilized, and such

data could be used to formulate approximations of catchments of specific sites. Use of ethnographic or ethnohistoric data should also be useful in

some areas.

Once an approximation of a catchment is made, the kinds of resources

126 DONNA C. ROPER

evaluated, the way they are analyzed, and the detail to which they are

analyzed also vary. Most analyses employ general land classifications of

some kind. For example, Vita-Finzi and Higgs (1970) used a series of

"land use capability classes" which are based on modern land uses.

These include irrigated land, arable, rough grazing, good grazing/

potentially arable, seasonal marsh, sand dunes, and irrigated crops

(Vita-Finzi and Higgs 1970:17). Within the time contours drawn around

each site, they evaluated the acreage of the enclosed territory and the

percentage of it occupied by each land type. Many other studies have

been similarly performed. Some analyses are explicitly based largely on one kind of resource,

such as soil or vegetation. For example, Webley's (1972) analysis of Tell

Gezer and several other sites in Palestine was based entirely on analysis of soils and their potential productivity. Peebles' (n.d.) study of Mound

ville, Alabama, was similarly conceived. Adams (1977) was largely con

cerned with plant potential, and Roper (1974) also relied heavily on floral

zones.

Two problems may arise with resource availability estimation: (a) Should inference about important factors in site location be based on a

site's relation to a single type of resource? (b) How reliable are the recent

or modern distributions? Even the simplest models of site location specify location as being

determined by the interaction of several variables. For example, Chisholm (1968:102-103) has listed water, arable land, grazing land, fuel, and building materials as "the five basic elements of ... a settler com

munity's economy: with none can the settlement dispense." Jochim

(1976:50) has listed as primary goals in settlement placement among

hunter-gatherers: "1. Proximity of economic resources. 2. Shelter and

protection from the elements. 3. View from observation of game and strang ers." Hill (1971:56) has diagrammed a multivariable model of the deter

minants of site locations, including critical resources, their proximity and

spacing, population density, and other variables. The use of single re

source types such as soil or vegetation may therefore be unfairly limiting. It may not necessarily allow inference about why a site is located where it

is or how it may have functioned in a settlement system. It does, of

course, permit description of how sites are located relative to soils or

vegetation, and a comparison among sites of their potential for certain

economic activities. If this is the goal of a specific study, then use of only a few resource types is fine. More complete modeling of settlement

location and the settlement system, however, requires the use of a wider

variety of resource types. Site catchment analysis would be virtually impossible if it were based

SITE CATCHMENT ANALYSIS 127

on anything but modern or recent resource distributions, since maps of

past resource distributions are seldom available. However, for a variety of reasons, including geomorphic change, climatic change, fluctuations in

sea level, and drastic changes in resource distribution with the introduc

tion of modern land use practices, modern data may be highly unreliable.

The problem cannot be adequately treated here, except to say that it

will be necessary to evaluate the case for each area. For example, Higgs and Vita-Finzi (1966:28) refer specifically to changing potentials for ex

ploitation of certain areas in Epirus, Greece. Vita-Finzi's (1969b) study, The Mediterranean Valleys is a consideration of how the streams feeding the Mediterranean have modified their courses in the last two millenia

(Vita-Finzi 1969b: 1) and concludes with a summary of the implications of

geomorphic change for available exploitable area (Vita-Finzi 1969b: 118). Adams (1977) employs a "grazing filter" for attempting to assess changes in plant food potential surrounding sites in the Rio Puerco Valley in New

Mexico. Many archaeologists in the Midwestern United States have used

(for site catchment analysis or other purposes) vegetation reconstructions

based on the Government Land Office surveys. These were done in the

public land states of the United States at about the time of Euro-American

settlement. Although they provide reasonable estimates of major vegeta tion zones, they must be used with caution, for they too were done under

varying climatic conditions and other conditions that may affect their

utility for prehistoric resource estimation (Wood 1976; King 1978). In

some places, and for some time periods, their use could be completely

misleading. The techniques for analysis of site catchment data also vary. Many

studies have evaluated the data by inspection of tables or drawings of resource zones surrounding sites (see Barker 1975b for the most exhaus

tive example of this approach). Sometimes, particularly if a number of

sites are being evaluated, this interpretation is graphically assisted with

pie diagrams (e.g., Vita-Finzi and Higgs 1970) or histograms (e.g., Ellison

and Harriss 1972; Barker 1972) of land type proportions. Roper (1974,

1975) and Baumler (1976) both used multivariate statistical techniques (factor analysis, multidimensional scaling, and cluster analysis) for de

scribing and comparing site territories and their resource potential.

Flannery (1976b:92-93) has suggested, however, that the relevant ques tion may indeed not be what percent of a particular land use type falls

within the territory, but rather whether or not this is sufficient land for the

needs of the site's inhabitants and whether or not it is significantly more

than could be expected by chance. The answer to the former question, he

says, requires estimation of both site population and available land.

Rossman (1976:102), Flannery (1976a: 107), and Zarky (1976:122) all ad

128 DONNA C. ROPER

dress such estimates in their analyses of site catchments in Formative

Mesoamerica. The latter question requires the use of inferential statistics

for an answer. Zarky's (1976) analysis of site catchments at Oc?s in

Guatemala used percentage point differences, chi-squared tests, and

binomial tests for evaluating which environmental zones were repre sented in higher proportions immediately surrounding a site than they

were in the total study area.

Common to many site catchment studies is the evaluation of all land

types as if they were of equal value for what they produce. This simply is

not true, however. Seasonal and spatial disparities in potential is one of

the reasons site catchment analysis was originally developed, but few

studies actually quantify this potential. A number of exceptions should be

discussed.

The simplest and oldest means of accounting for differentials is that

used by Vita-Finzi and Higgs (1970:30). In their study, land at greater distances from agricultural sites was weighted proportionally less than

was that close to the site, to compensate for the increased travel time

required. The area within 1 km was weighted 100%; 1-2 km, 50%; 2-3

km, 33%; 3-4 km, 25%; and 4-5 km, 20%; and the figures were tabled and

graphed accordingly (Vita-Finzi and Higgs 1970:28-31). Rossman

(1976:100-101) used these same weighting figures in his analysis of exploi tation territories in the San Lorenzo area in Mesoamerica. Cassels

(1972a: 209) adjusted them only slightly (and only because he used a

different set of radii) for his analysis in the Waikato area of New Zealand.

Flannery's (1976b) suggestion that population and yield estimates may be used to evaluate the sufficiency of land to support the population has

already been mentioned. A number of other studies have been even more

specific about producing some estimate of yield of the resource zones

within the catchment rings of the sites, without necessarily relating the

figures to population density. Peebles' (1978) study of Moundville Phase

settlement in Alabama measured areas of different soil types within 1 and

2 km of Moundville Phase sites, and estimated "gross median produc

tivity" of each ring. This was done by multiplying "the midpoints of the

range of the average yields of bushels of corn for each soil type in the

catchment" and summing the products (Peebles 1978). Webley (1972:178),

using soil types, similarly estimated barley yields and goat potentials for a

single site in Palestine.

In all these cases, however, estimates have been made for only one or

two resources?albeit important ones. Two studies have attempted to

quantify a larger number of resources and to incorporate the seasonally different potential as well. Cassels (1972a:209-213) corrected not only for

distance from a site, but also for seasonal difference in potential of each

zone.

SITE CATCHMENT ANALYSIS 129

Adams (1977) attempted to assess differential wild plant resource po tential for five sites in the Rio Puerco Valley of New Mexico. She de

veloped a set of scores using parts of a plant, the season or seasons of the

year when the plant is available, its dependability, and a rough estimate of

the work necessary for its procurement and preparation. Seventy-five

plants scored included those documented ethnographically for the South

west and archaeologically for the Rio Puerco. Plant scores were then

summed to determine a total tally for the land lying within an arbitrary radius of each site. This technique has the obvious advantage of estimat

ing potential only for those plants most likely employed by the. inhabitants

of the area. Although seasonal considerations enter the calculation of

plant scores, they do not produce an estimate of seasonal fluctuations in

resource potential, as does Cassels' (1972a) analysis; rather, they estimate

the total annual potential within the arbitrary radius of each site.

Adams' study is not unlike a study by Munson et al. (1971) in which the

yields of several plant and animal species of major economic importance within 1.78 miles (10 square miles) of the Scovill site in Illinois were

estimated and the figures compared with actual remains to evaluate

whether or not animals and plants were being taken in proportion to their

availability. Munson and associates conclude that there was no selection

for plant species, but that there was a selection for animal species (Mun son et al. 1971:426). An alternative explanation, however, could simply be

that plants were collected from the immediate environs of the site, whereas animals were not. In other words, if it is assumed that the

samples were good, the observed disparities could simply be a result of zonation of exploitation of resources.

Two studies stand apart from most site catchment studies in not confin

ing themselves to a small, circumscribed area surrounding a site. Foley

(1977) developed an ecological model accounting for differential produc

tivity in an area. This model was free of specific loci, instead using

quadrats superimposed on a general resource zone map of an area in

which some sites were assumed to be located. This approach would then

analyze the energy balance by subtracting the value of the energy neces

sary to exploit an area from a given locus from the extracted energy

(Foley 1977:164-165, 177-181). The approach is interesting, but unfortu

nately it is illustrated with a hypothetical example.

Flannery's (1976a) study, "Empirical Determination of Site Catch ments in Oaxaca and Tehuac?n," although tied to specific sites, is

likewise free of an arbitrary analytic territory. Instead of evaluating what

resources were available to a site's inhabitants within some arbitrarily (and perhaps unrealistically) demarcated area, Flannery reversed the

procedure, started with data on the plant, animal, and mineral resources

found at sites, and asked, "From how far away must they have come?"

130 DONNA C. ROPER

(Flannery 1976a: 103). The analysis considered all kinds of resources, from the commonest plants to the most exotic trade items. It obviously avoided the problem of arbitrarily superimposing a territory; it also re

quired good faunal and floral preservation, detailed study of those re

mains, and comprehensive knowledge of resource distributions. Flan

nery's conclusions expectably suggest a zonation of resource use

(1976a: 117), but also document a total catchment area/ar larger than the

analytic territory used in most studies. Most basic plant and mineral needs

were satisfied within 5 km of the site, but animals, wood, and exotic

materials came from farther away. It is still a documentation of the fact

that resources reflected at a site may come from a far broader area than

the small analytic territories used by most site catchment analyses. With the exception of the Foley (1977) and Flannery (1976a) papers,

therefore, procedures for site catchment analysis can be summarized as

follows. First, define the analytic territory. To do so, use a circle or circles

centered on the site or an irregularly shaped territory defined by time

contours, or infer the territory by the site's relation to its neighbors. If the

latter is done, be prepared to justify the assumption of site contem

poraneity. Then, measure the area of each resource zone within each

site's territory. Table these figures, or graph them, or use them in a

statistical analysis of site territories. Differential weighting of more distant

resources, estimates of yields, and accounting for differential seasonal

potentials may be used at this point. The exact procedure chosen and the

use made of the results of the analysis will depend on the purpose of the

analysis. It is to the purpose of site catchment analysis, therefore, that we

now turn.

WHY HAS SITE CATCHMENT ANALYSIS BEEN DONE?

The literature on site catchment analysis contains numerous contribu

tions from both British and American researchers. To assess the diversity of site catchment applications, however, it is necessary to appreciate the

fact that it is set within quite different contexts in Europe and the

Americas.

European Studies

British archaeology has often been seen as largely concerned with

artifacts and historical reconstructions of their development. In the late

1960s, however, a new theme entered the British practice of arch

aeology?that cultural phenomena were not be be explained by cul

SITE CATCHMENT ANALYSIS 131

ture alone, but also by other factors, including economy. Higgs and

Jarman stated this theme quite explicitly at the conclusion of their essay

reconsidering the origins of agriculture:

The cultural model has dominated thought and speculation in European archaeology for

many decades .... The study of economy, the major selective force in prehistory

has, until now, largely been ignored. With its development, ... the whims, fashions,

and freedom of choice associated with cultures may become of less importance to

archaeology than the study in man's past of natural mechanisms as the true causes of

human behavior [1969:40].

The study of economies has long had a place in British archaeology,

particularly following Clark's (1952) major synthesis of the economy of

prehistoric Europe. However:

In spite of Clark's admonition that many archaeologists were too artefactually

oriented, the field has continued to be dominated by the consideration of artefactual

types and their chronology, and economic concepts have usually served for little more

than the classification of cultures into 'hunter-gatherers,' 'pastoralists,' or 'farmers'

[Higgs and Vita-Finzi 1972:27].

A major innovation in site catchment analysis was therefore the role

given to economy in interpretation of the archaeological record. Vita

Finzi and Higgs (1970:4) are quite clear in stating that attempts to solve

the problem of the nature of prehistoric economies "must take into

account not only the artefacts but also the possibilities inherent in the site

situations themselves." Their study of sites in the Mt. Carmel area of

Palestine, as well as the earlier study by Higgs et al. (1967) in Epirus,

Greece, were attempts to generate testable hypotheses about prehistoric economies from this perspective.

Several other studies had as their primary purpose to reconsider

culture-historical reconstructions from the perspective of economies, and

in so doing generated hypotheses about those economies by using site

catchment analysis. The studies by Graeme Barker (1972, 1973, 1975b) on

the Bronze Age of central Italy are a case in point. Barker's analysis

sought to refute the standard concept of central Italian prehistory, which

postulates "a series of neat cultural phases, each with its economic label,

developing with regional variations one after the other" (Barker

1973:359-360), by using, among other data, "the evidence of the location

of the sites themselves" (Barker 1973:360). Consideration of the ter

ritories of a series of sites (Barker 1972:198, Fig. 10) and their productive

potentials led to what Barker (1972:189) described as "a hypothetical reconstruction of related economic networks of sites, within an ecological rather than a cultural framework," as well as to an apparent major revision of at least the "economic labels" of the "neat cultural phases" that characterized the study of central Italian prehistory. The same au

132 DONNA C. ROPER

thor's study of early Neolithic sites in Yugoslavia (Barker 1975a), and the

study by Dennell and Webley (1975) of Neolithic and Bronze Age sites in

southern Bulgaria, both similarly proposed alternative settlement system models to the traditional interpretations, using site catchment analysis to

interpret the relation between site location and potential for occupation. Other studies were less concerned with alternatives to culture

historical scenarios, instead seeking to examine the economic aspects of

various cultural units: Webley's (1972) study of Tell Gezer and nearby sites in Palestine; Jarman and Webley's (1975) study of prehistoric sites in

Capitanata, Italy; Davidson's (1976) study of several Paleolithic sites in

Spain; Jarman's (1972, 1976) examination of prehistoric sites in Italy (the 1972 paper is actually more concerned with presentation of a model than

with the substantive results of the analysis); and Ellison and Harriss'

(1972) study of prehistoric and early historical sites in southern England all belong to this genre of site catchment studies. This is not to suggest that all these studies are similarly conceived or executed, however. Some

rely primarily on site location data (e.g., Webley 1972; Jarman and Web

ley 1975), whereas others augment the locational data with analysis of

faunal remains (e.g., Jarman 1976; Davidson 1976). Ellison and Harriss

(1972) were "concerned primarily with the locations of individual sites

and with what can be inferred from them" (Ellison and Harriss 1972:913) rather than with reconstruction of economies. To this end, site catchment

analysis was the primary line of evidence used, but it was supplemented

by other locational techniques for study of land use.

The final major purpose for which site catchment studies have been

performed in Europe has been the examination of the environmental context of single sites. For these studies, the term site catchment analysis is hardly appropriate, for neither are they analyses nor are they concerned

with catchments in the proper sense of the term. Rather, they are site

reports in which the site is related to its natural setting by description of

the area within a 5-km (or whatever) radius of the site?that is, the area

presumed to provide the majority of resources to the site. Details of the

surrounding territory are frequently shown in a drawing and briefly de

scribed. Reports by Noy et al. (1973) on Nahal Oren, Israel; by Moore et

al. (1975) on Tell Abu Hureyra, Syria; by Fagan (1976) on Gwisho,

Zambia; and by Clark (1972) on his reexamination of Star Carr, England, all fall in this category. Clarke's (1972) examination of Glastonbury,

England, is similar; here the site exploitation territory was considered in

building a model of the Iron Age society represented at the site.

In spite of the explicit concern for the study of the economy, and in

spite of the frank regard for behavior and development of laws of human

behavior (e.g., Higgs and Jarman 1975:2), many European studies are

SITE CATCHMENT ANALYSIS 133

directed toward clarification of what happened in prehistory rather than

with the development of such laws. In all these studies, the assumptions of site catchment analysis and the validity of the distance figures given by Lee, Chisholm, and others are taken as givens, and never have they been

subjected to confirmation or disconfirmation. However, Lee's (1969)

figures on the distance agriculturalists are willing to travel to tend their

fields may be applicable to the societies on which they are based, but they have not yet been shown to have universal validity. It is this almost

mechanical use of these figures for delineating the analytic territory that was noted earlier as one of the major problems with site catchment

analysis.

American Studies

Site catchment analysis is only beginning to be used by American

archaeologists. Some of the analyses in the Americas (especially Rossman

1976; Zarky 1976) have been performed for purposes similar to those of some of the British studies?that is, to examine and quantify the nature of

the territory immediately accessible to a site's inhabitants. American

archaeology has, however, had a strong tradition of settlement pattern

analysis. Beginning with Gordon R. Willey's (1953) studies in the Vir?

Valley of Peru and continuing up to the present, the settlement pattern

concept has been a functional rather than a historical concept. The ar

chaeologist interested in settlement patterns defined a series of site types which can be used for a variety of purposes. One of these is the study of their distribution across the landscape and the explanation of location.

Site catchment analysis has been used in American archaeology primarily for modeling the spatial distribution of functionally distinct sites within a

settlement system, or for examination of the resource potential of sites

thought to have occupied different positions in a settlement system.

Roper's (1975) study of settlement patterns of Woodland sites in central Illinois is an example of the former use of site catchment analysis. A

series of very generally defined site types was recorded during a survey of

the Sangamon River Valley, and the site catchment data were used to

assess the locations of each site. After a statistical analysis was made of

these data, Roper was able to postulate a general, but testable, model of

Middle and Late Woodland settlement patterns in the valley, and to

provide a general test of this model using such limited excavation data as

were available, survey data not used in formulation of the site types, and

comparative literature.

An example of the latter use is Peebles' (1978) study of Moundville Phase settlement in Alabama. Peebles used soils and their estimated

134 DONNA C. ROPER

yields to examine resource potential at functionally different sites in the

Moundville Phase. His techniques have already been described. Addi

tionally, he related estimated productivity to site size, assuming that there

is a relationship between population size and subsistence base, between

settlement size and resident population, and between area of scatter and

settlement size. The hypothesis was that if a prime criterion for location is

agricultural land, then site size should vary with soil productivity. The

Pearson product-moment correlation coefficient was used to test the

hypothesis. Village-hamlet settlements did show a strong correlation

between size and productivity, while the minor ceremonial centers had a

rather low relationship (Peebles 1978). Brumfiel's (1976) test of a population pressure hypothesis in a part of

the Valley of Mexico was similarly conceived and executed. The basic

argument was expressed as follows:

We could expect a situation of population pressure to be expressed in a simple correlation between the relative number of inhabitants at each village and the relative

productive potential of agricultural land available at each village [Brumfiel 1976:237,

italics in the original].

Brumfiel implicitly made the same assumption as did Peebles but used

linear regression to predict site size from productive potential?a measure

derived from data on available agricultural land and the fertility of that

land (Brumfiel 1976:240). Browman (1976) used the premises of site catchment analysis as ex

pressed by Higgs and associates (Higgs and Vita-Finzi 1972; Jarman 1972; Jarman et al. 1972) to derive predictions about expected spacings of

several site types. These predictions are then compared with actual

spacings of sites (Browman 1976:471). Using an estimated biomass for the

catchment, he then postulated the demographic processes operative in the

Jauja-Huancayo basin of Peru at around a.d. 500 (Browman 1976:473

474). This use of the site catchment model, while holding in common with

Peebles' and Brumfiel's the incorporation into demographic studies, is the

only analysis in which the assumptions of site catchment analysis were

used to generate predictions about the form of the archaeological record, and the only study that could be construed to be a test of the validity of

those assumptions.

Finally, we might note Hassan's (1975) use of the catchment concept. Hassan's paper is a general discussion of population density, size, and

growth rate. In developing the discussion of size, Hassan (1975:38) notes

that "population size is a function of the population density and area."

Incorporating the catchment concept as a measure of area, he then

(1975:39-40) uses it to discuss the relationships between size and density.

SITE CATCHMENT ANALYSIS 135

It is quite clear that the relationship is variable, suggesting that the

attempt to use similar-sized analytic territories in different kinds of places is not only an unrealistic way to proceed in general but is subject to

varying degrees of error.

CONCLUSIONS

It has been a major purpose of this chapter to summarize and review

the assumptions of site catchment analysis, the techniques employed in its

implementation, and the purposes for which it has been performed. It

should be clear that there is a wide diversity in purpose and technique. The criterion for selection of papers discussed here was that the authors

said they were using site catchment analysis or something resembling site

catchment analysis. It remains then to come to a conclusion as to exactly what is site catchment analysis.

Excluding those reports using a circular territory around a site solely for environmental description purposes, consideration of the remainder of the studies suggests that site catchment analysis is most correctly viewed as a method?that is, a set of techniques for analysis of data (Dunnell 1971:34). Its unity derives from the incorporation of techniques relating site location to resource availability within the territory immediately sur

rounding a site. It assumes that this territory, which will be of relatively finite size and show seasonal variability, is of primary importance in

provisioning the site's residents. How this territory is determined for

analysis, however, will depend on whether it is preferable to use an

arbitrarily demarcated area (based on either time or distance contours) or to attempt to approximate the area from spacing of the sites. The latter

requires slightly more stringent assumptions about the sites analyzed, but

provides a better estimation of the behavioral unit (that is, the catchment

itself) if it can be shown to be appropriate. The kind of resources analyzed (soil, topography, flora, etc.) will depend on availability of data and the researcher's beliefs about what kinds of resources were important to the

community under study. Site catchment analysis is therefore useful with a

variety of locational models.

Further, it can be, and has been, used in studies of a variety of types

including evaluation of the feasibility of various culture-historical recon

structions (e.g., Barker 1972, 1973, 1975a,b), determination of the feasibil

ity of various forms of economy (e.g., Higgs et al. 1967r; Vita-Finzi and

Higgs 1970), modeling settlement patterns (e.g., Peebles 1978; Roper 1975), and study of demographic processes (e.g., Browman 1976;

Brumfiel 1976).

136 DONNA C. ROPER

Development of site catchment analysis is, of course, not without

problems. These are discussed in the text of this paper. It is now an infamous fact that archaeologists are prone to borrow

methods rather than develop their own. Yet it has also been pointed out

by a number of authors that the tendency to do so ignores archaeology's greatest asset?the time depth unavailable to most other social

scientists?and it ignores the possibility that the presently known range of cultural expressions does not represent a full range of all societies that

have ever existed. The development of site catchment, although it draws

on ethnography and geography, has been developed with the considera

tion of the potentials of the archaeological record. It is to be hoped that

the needed development and refinement of the method will occur.

ACKNOWLEDGMENTS

This chapter in its several versions has benefited considerably from the comments of Karen

R. Adams, Vorsila L. Bohrer, Sharon L. Brock, Carole L. Crumley, Susan K. Goldberg, Michael B. Schiffer, V. Ann Tippitt, and, of course, the ever-helpful Anonymous Review

ers. Dr. Bohrer and Ms. Adams were kind enough to provide me with copies of several of

their unpublished papers as well as to correct my summary of their approach; Christopher S. Peebles provided an advance copy of his Moundville manuscript. Special thanks go to

Michael B. Schiffer for inviting me to write this paper. The result ofthat invitation was not

only the opportunity but the necessity to assess more critically a method which I, too, had

perhaps rather blindly accepted for several years. My own analyses will never be the same

again. The traditions of scholarship require that I absolve all the above-named from liability for remaining errors and shortcomings?a requirement with which I am most happy to

comply. Several very capable colleagues have my undying appreciation for keeping things going,

but keeping me involved, informed, and sane, while I completed the final draft of this chapter

during the launching of a major field season.

REFERENCES

Adams, K. R.

1977 Site catchment analysis of wild plant resources in the heavily overgrazed Rio

Puerco Valley of New Mexico. Paper presented at the 42nd annual meeting of

the Society for American Archaeology, New Orleans.

Barker, G.

1972 The conditions of cultural and economic growth in the Bronze Age of Central

Italy. Proceedings of the Prehistoric Society 38:170-208.

1973 Cultural and economic change in the prehistory of Central Italy. In The explana tion of culture change: Models in prehistory, edited by C. Renfrew. Pittsburgh:

University of Pittsburgh Press. Pp. 359-370.

1975a Early neolithic land use in Yugoslavia. Proceedings of the Prehistoric Society 41:85-104.

SITE CATCHMENT ANALYSIS 137

1975b Prehistoric territories and economies in Central Italy. In Palaeoeconomy,

edited by E. S. Higgs. London and New York: Cambridge University Press. Pp.

111-175.

Baumler, M.

1976 An initial prehistoric settlement-subsistence analysis for the Little Blue River

Area (Little Blue Channel-Modification Project, Archaeological Research De

sign. Report to the U.S. Army Corps of Engineers). Lawrence: Department of

Anthropology, University of Kansas. Pp. 11-55.

Browman, D. L.

1976 Demographic correlations of the Wari conquest of Junin. American Antiquity

41(No. 4):465-477.

Brumfiel, E.

1976 Regional growth in the Eastern Valley of Mexico: A test of the "population

pressure" hypothesis. The early Mesoamerican village, edited by K. V. Flan

nery. New York: Academic Press. Pp. 234-249.

Cassels, R.

1972a Human ecology in the prehistoric Waikato. The Journal of the Polynesian

Society 81:196-247.

1972b Locational analysis of prehistoric settlement in New Zealand. Mankind 8:212

222.

Chisholm, M. C.

1968 Rural settlement and land use: An essay in location. Chicago: Aldine.

Clark, J. G. D.

1952 Prehistoric Europe: The economic basis. New York: Philosophical Library. 1972 Star Carr: A case study in bioarchaeology. Addison-Wesley Modular Publica

tions in Anthropology No. 10.

Clarke, D. L.

1968 Analytical archaeology. London: Methuen.

1972 A provisional model of an Iron Age society and its settlement system. In Models

in archaeology, edited by D. L. Clarke. London: Methuen. Pp. 801-869.

1977a Spatial information in archaeology. In Spatial archaeology, edited by D. L.

Clarke. New York: Academic Press. Pp. 1-34.

1977b (editor) Spatial archaeology. New York: Academic Press.

Davidson, I.

1976 Les Mallaetes and Mond?ver: The economy of a human group in prehistoric

Spain. In Problems in economic and social archaeology, edited by G. de G.

Sieveking, I. H. Longworth, and K. E. Wilson. Boulder: Westview Press. Pp. 483-499.

Dennell, R. W., and D. Webley 1975 Prehistoric settlement and land use in southern Bulgaria. In Palaeoeconomy,

edited by E. S. Higgs. London and New York: Cambridge University Press. Pp. 97-109.

Dunnell, R. C.

1971 Systematics in prehistory. New York: Free Press.

Ellison, A., and J. Harriss

1972 Settlement and land use in the prehistory and early history of southern England: A study based on locational models. In Models in archaeology, edited by D. L.

Clarke. London: Methuen. Pp. 911-962.

Fagan, B. M.

1976 The hunters of Gwisho: A retrospect. In Problems in economic and social

138 DONNA C. ROPER

archaeology, edited by G. de G. Sieveking, I. H. Longworth, and K. E. Wilson.

Boulder: Westview Press. Pp. 15-24.

Findlow, F. J., and S. P. DeAtley

1974 Prehistoric land use patterns in the Animas Valley: A first approximation.

Anthropology UCLA 6(No. 2): 1-57.

Flannery, K. V.

1976a Empirical determination of site catchments in Oaxaca and Tehuacan. In The

early Mesoamerican village, edited by K. V. Flannery. New York: Academic

Press. Pp. 103-117.

1976b The village and its catchment area: Introduction. In The early Mesoamerican

village, edited by K. V. Flannery. New York: Academic Press. Pp. 91-95.

Foley, R.

1977 Space and energy: A method for analysing habitat value and utilization in

relation to archaeological sites. In Spatial archaeology, edited by D. L. Clarke.

New York: Academic Press. Pp. 163-187.

Haggett, P.

1965 Locational analysis in geography. New York: St. Martin's.

Hamilton, W. J. Ill, and K. E. F. Watt

1970 Refuging. Annual Review of Ecology and Systematics 1:263-286.

Hassan, F. A.

1975 Determination of the size, density, and growth rate of hunting-gathering popu

lations. In Population, ecology, and social evolution, edited by S. Polgar. The

Hague: Mouton. Pp. 27-52.

Higgs, E. S. (editor) 1975 Palaeoeconomy. London and New York: Cambridge University Press.

Higgs, E. S., and M. R. Jarman

1969 The origins of agriculture: A reconsideration. Antiquity 43:31-41.

1975 Palaeoeconomy. In Palaeoeconomy, edited by E. S. Higgs. London and New

York: Cambridge University Press. Pp. 1-7.

Higgs, E. S., and C. Vita-Finzi

1966 The climate, environment, and industries of Stone Age Greece: Part II. Pro

ceedings of the Prehistoric Society 32:1-29.

1972 Prehistoric economies: A territorial approach. In Papers in Economic Prehis

tory, edited by E. S. Higgs. London and New York: Cambridge University Press. Pp. 27-36.

Higgs, E. S., C. Vita-Finzi, D. R. Harriss, and A. E. Fagg 1967 The climate, environment, and industries of Stone Age Greece: Part III. Pro

ceedings of the Prehistoric Society 33:1-29.

Higgs, E. S., and D. Webley 1971 Further information concerning the environment of Palaeolithic Man in Epirus.

Proceedings of the Prehistoric Society 37(Part II):367-380.

Hill, J. N.

1971 Research propositions for consideration. Southwestern anthropological re

search group. In The Distribution of Prehistoric Population Aggregates, edited

by G. J. Gumerman. Prescott College Anthropological Papers Arizona: Pres

cott College Press. (No. l):55-62.

Hodder, I. R.

1977 Some new directions in the spatial analysis of archaeological data at the regional scale. In Spatial archaeology, edited by D. L. Clarke, New York: Academic

Press. Pp. 223-351.

SITE CATCHMENT ANALYSIS 139

Hodder, I. R. and C. Orton

1976 Spatial analysis in archaeology. London and New York: Cambridge University

Press.

Jarman, M. R.

1972 A territorial model for archaeology: A behavioral and geographical approach. In

Models in Archaeology, edited by D. L. Clarke. London: Methuen. Pp. 705

733.

1976 Prehistoric economic development in sub-alpine Italy. In Problems in economic

and social archaeology, edited by G. de G. Sieveking, I. H. Longworth, and K.

E. Wilson. Boulder: Westview Press. Pp. 523-548.

Jarman, M. R., C. Vita-Finzi, and E. S. Higgs 1972 Site catchment analysis in archaeology. In Man, settlement, and urbanism,

edited by P. J. Ucko, R. Tringham, and G. W. Dimbleby. Cambridge, Mass:

Schenkman. Pp. 61-66.

Jarman, M. R., and D. Webley

1975 Settlement and land use in Capitanata, Italy. In Palaeoconomy, edited by E. S.

Higgs. London and New York: Cambridge University Press. Pp. 177-221.

Jochim, M. A.

1976 Hunter-gatherer subsistence and settlement: A predictive model. New York:

Academic Press.

King, F. B.'

1978 Additional cautions on the use of the GLO survey records in vegetational

reconstruction in the Midwest. American Antiquity 43(No. 1):99?103.

Kroeber, A. L.

1939 Cultural and natural areas of native North America. University of California

Publications in Archaeology and Ethnology No. 39

Lee, R. B.

1969 !Kung Bushman subsistence: An input-output analysis. In Environment and

cultural behavior, edited by A. P. Vayda. Garden City, N. Y.: Natural History Press. Pp. 47-79.

Moore, A. M. T., G. C. Hillman, and A. J. Legge 1975 The excavation of Tell Abu Hureyra in Syria: A preliminary report. Proceed

ings of the Prehistoric Society 41:50-77.

Munson, P. J., P. W. Parmalee, and R. A. Yarnell

1971 Subsistence ecology of Scovill, a terminal Middle Woodland village. American

Antiquity 36(No. 4):410-431.

Noy, T., A. S. Legge, and E. S. Higgs 1973 Recent excavations at Nahal Oren, Israel. Proceedings of the Prehistoric Soci

ety 39:75-99.

Peebles, C. S.

1978 Determinants of settlement size and location in the Moundville Phase. In

Mississippian settlement patterns, edited by B. D. Smith. New York: Academic

Press. Pp. 369-416.

Roper, D. C.

1974 The distribution of Middle Woodland sites within the environment of the Lower

Sangamon River, Illinois. Illinois State Museum Reports of Investigations No.

30.

1975 Archaeological survey and settlement pattern models in central Illinois. Unpub lished Ph.D. dissertation, University of Missouri-Columbia.

140 DONNA C. ROPER

Rossman, D. L.

1976 A site catchment analysis of San Lorenzo, Veracruz. In The early Mesoameri

can village, edited by K. V. Flannery. New York: Academic Press. Pp. 95-103.

Schiffer, M. B.

1975 An alternative to Morse's Dalton settlement pattern hypothesis. Plains An

thropologist 2<KNo. 70):253-266.

Steward, J. H.

1938 Basin-plateau aboriginal socio-political groups. Bureau of American Ethnology

Bulletin No. 120.

Vita-Finzi, C.

1969a Early man and environment. In Trends in geography?An introductory survey,

edited by U. Cooke and J. H. Johnson. Oxford: Pergamon. Pp. 102-108.

1969b The Mediterranean valleys. London and New York: Cambridge University Press.

Vita-Finzi, C, and E. S. Higgs, 1970 Prehistoric economy in the Mount Carmel area of Palestine: Site catchment

analysis. Proceedings of the Prehistoric Society 36:1-37.

Webley, D.

1972 Soils and site location in prehistoric Palestine. In Papers in economic prehis

tory, edited by E. S. Higgs. London and New York: Cambridge University Press. Pp. 169-180.

Willey, G. R.

1953 Prehistoric settlement patterns in the Viru Valley, Peru. Bureau of American

Ethnology Bulletin No. 155.

Willey, G. R., and P. Phillips 1958 Method and theory in American Archaeology. Chicago: University of Chicago

Press.

Wilmsen, E. N.

1973 Interaction, spacing behavior, and the organization of hunting bands. Journal of

Anthropological Research 29(No. 1): 11-31.

Wood, W. R.

1976 Vegetational reconstruction and climatic episodes. American Antiquity 41(No.

2):206-208.

Zarky, A.

1976 Statistical analysis of site catchments at Ocos, Guatemala. In The early

Mesoamerican village, edited by K. V. Flannery. New York: Academic Press.

Pp. 117-128.